Respiratory Microbiota Associations with Asthma Across American and Emirati Adults: A Comparative Analysis

Abstract

Background: Clinical features of asthma are associated with differences in the lower airway microbiome. However, knowledge is limited on whether airway microbiota composition differs between individuals residing in different geographic regions and if asthma-associated differences in lower airway microbiota are similar between distinct populations.

Methods: Existing 16S rRNA gene sequence data, generated from sputum collected from adults with or without asthma (n = 74) from two single-center cohort studies in the U.S. and United Arab Emirates, were re-processed for merged computational analysis using standard available tools. Potential differences between study sites, asthma status and specific clinical factors (inhaled corticosteroid use, ICS; obesity) were examined.

Results: Differences in sputum bacterial composition, assessed by alpha- and beta-diversity measures, were associated with study site. Despite this, asthma-related differences were discerned in both cohorts. Specifically, sputum microbiota of asthmatic patients on ICS treatment displayed reduced bacterial phylogenetic diversity, compared to those not on ICS treatment (p = 0.006). Sputum bacterial composition also differed by obesity status (unweighted Unifrac distance PERMANOVA, p = 0.004). Specific genera were identified in both cohorts that were differentially enriched between obese vs. non-obese subjects, including Rothia and Veillonella (obesity-associated) and Campylobacter (non-obesity-associated).

Conclusions: Our findings suggest clinical factors associated with differences in the lower airway microbiome in asthma may transcend variation related to geographic area of residence.

Keywords: airway; asthma; geography; microbiome.

Figure A1.

LefSe analysis of U.A.E. sputum bacterial microbiota data showing bacterial genera identified as differentially abundant between asthmatic patients on inhaled corticosteroid treatments (ACS), those not on inhaled steroids (ATH), and healthy controls. (LDA score > 2.0, p < 0.05).

Figure 1.

(A,B) Sputum bacterial alpha-diversity between study sites (Michigan (n = 55) and United Arab Emirates, U.A.E. (n = 19) * p ≤ 0.05; ** p ≤ 0.001. (C) Principal coordinates analysis based on beta-diversity measures (Bray–Curtis distance, p = 0.008; Unifrac distances, p = 0.001) showing variation in overall sputum bacterial composition associated with study site.

Figure 2.

Differentially abundant bacterial genera between the study sites by LefSe analysis. LDA score > ∣2.5∣, padj ≤ 0.05). Positive LDA score values indicate enrichment in the MI group. Negative LDA scores conversely indicate enrichment in the UAE group.

Figure 3.

(A) Sputum bacterial phylogenetic diversity is lowest in the asthmatic group on inhaled corticosteroid treatment (ACS, n = 30), compared to those not on inhaled steroids (ATH, n = 24, p = 0.006) or healthy controls (CON, n = 20, p = 0.03). (B) Principal coordinates analysis by Bray–Curtis (left) or unweighted Unifrac (right) distance, showing the variation in overall sputum bacterial composition, largely driven by the ACS group (p = 0.008). Bray–Curtis distance PERMANOVA, (p = 0.008), unweighted Unifrac distance PERMANOVA, (p = 0.013). ATY, atopic non-asthma control (MI cohort only).

Figure 4.

Principal coordinates analysis using unweighted Unifrac distance showing airway bacterial composition difference between obese (ob) and non-obese (nob) asthmatic subjects (PERMANOVA p = 0.004) across both cohorts. Differentially abundant bacterial genera between nob and ob asthma identified by LefSe (LDA score > ∣3.0∣, FDR < 0.05). Positive LDA scores indicate enrichment in the nob group.

Figure 5.

LefSe analysis of U.A.E. sputum microbiota data only demonstrating bacterial genera differentially abundant between obese (ob) and non-obese (nob) participants. Genera in bold text indicate those associated with nob and ob status in both cohorts.